Syntax command for dropping single cases

[daniel_wild]

Hello,

I hope someone could help me regarding single cases, which have been identified as outliers during regression analysis (st.residuals > 2; > 3).

Is there a practical syntax command for this. And if there is one where to include it? Simply adding at the end of commands for regression in the syntax?

Thanks in advance!

Kind regard

Daniel

[Englund]

What software do you use? In STATA you just write:

drop if residual>3

That code drops the whole row. I suppose that is what you want?!

[daniel_wild]

Thanks for the reply.

I use SPSS20. Actually that is not quiet what I mean.

I would like to drop the specific respondent or case. Because I dont want to save the variables for the residuals all the time and then analyze them seperately...that is quiet cumbersome if you use up to 10 predictors..

Kind regards

daniel